Vulnerabilities

Redash is a package for data visualization and sharing. From 5.0.2 to 26.3.0, the get_next_path() function in Redash's authentication module stripped the scheme and netloc from user-supplied next parameters but did not normalize multiple leading slashes, allowing a crafted login URL such as /login?next=////evil.com to redirect users to an external attacker-controlled site after authentication.

Redash is a package for data visualization and sharing. In versions 10.0 and priorm the implementation of URL-loading data sources like JSON, CSV, or Excel is vulnerable to advanced methods of Server Side Request Forgery (SSRF). These vulnerabilities are only exploitable on installations where a URL-loading data source is enabled. As of time of publication, the `master` and `release/10.x.x` branches address this by applying the Advocate library for making http requests instead of the requests library directly. Users should upgrade to version 10.0.1 to receive this patch. There are a few workarounds for mitigating the vulnerability without upgrading. One can disable the vulnerable data sources entirely, by adding the following env variable to one's configuration, making them unavailable inside the webapp. One can switch any data source of certain types (viewable in the GitHub Security Advisory) to be `View Only` for all groups on the Settings > Groups > Data Sources screen. For users unable to update an admin may modify Redash's configuration through environment variables to mitigate this issue. Depending on the version of Redash, an admin may also need to run a CLI command to re-encrypt some fields in the database. The `master` and `release/10.x.x` branches as of time of publication have removed the default value for `REDASH_COOKIE_SECRET`. All future releases will also require this to be set explicitly. For existing installations, one will need to ensure that explicit values are set for the `REDASH_COOKIE_SECRET` and `REDASH_SECRET_KEY `variables.

Redash 8.0.0 is affected by LDAP Injection. There is an information leak through the crafting of special queries, escaping the provided template since the username included in the search filter lacks sanitization.

Havoc Research discovered an authenticated Server-Side Request Forgery (SSRF) via the "JSON" data source of Redash open-source 8.0.0 and prior. Possibly, other connectors are affected. The SSRF is potent and provides a lot of flexibility in terms of being able to craft HTTP requests e.g., by adding headers, selecting any HTTP verb, etc.